iofiiiiiiiii

Yeah, when you have tall servers this can be a really surprising factor. In some sense you could view this as an extension of processor caching behaviors, which also causes some memory accesses to be lower - just due to cache behaviors, not physical location. But in many cases, the same tools can be used to fight both "far" memory accesses and cache trashing, by using a thread-isolated architecture.

I have been dealing with the topic for a few years now and it was surprisingly hard to track down the bottlenecks to actual numbers. Some time ago I managed to find a good example to demonstrate the effect in a tangible way and wrote up an article about it. If the topic sounds interesting, you might enjoy https://sander.saares.eu/2025/03/31/structural-changes-for-4... (Structural changes for +48-89% throughput in a Rust web service).

I used to run my personal website on Hugo but after a few years I wanted to upgrade the Hugo version and was suddenly out in the cold - there was no real path to upgrade, everything only works together once and as soon as I started upgrading, there was mismatch between the generator, the theme, and whatever widgets I used.

Moved to wordpress.com since that. No more worry about keeping things working, I can focus on the content. Admittedly, the horrible load times of wordpress.com sites are causing me to look at alternatives - waiting 5 seconds for the homepage to show up is not really acceptable.

I wish someone made a hosted version of a static site generator - they maintain the compatibility between individual components, provide some online editor for content, but the output is just a bunch of static files generated from this. Have not found one so far but if you know one, please drop a line!

Yeah, loved the Chinese series! It was very true to the book and exceeded my expectations!

What are the differences in the design principles of the AWS Rust SDK compared to AWS SDKs of other languages? In what ways is it special to work best with the Rust ecosystem?

What does your heart tell you? Palladium[1] came and went and then suddenly most laptops and mobile devices have a built-in TPM today. No doubt history will repeat.

[1] https://en.wikipedia.org/wiki/Next-Generation_Secure_Computi...

I recently saw people get rejected at the gate when they tried to board ahead of time in Frankfurt, so it sounds like they are starting to enforce it.

I once asked the developers about this and they did not seem to understand what the point of a TPM even is, unfortunately.

Agreed, I find myself having to think orthogonally to common sense whenever I try to use one of its SDKs. Nothing works the way you expect it to, everything has 3 layers of unnecessary abstraction and needs to be approached via the back door. Many features have caveats about when it works, where it works, how much it works, during what phase of the moon it works and how long your strings can be when Jupiter is visible in the sky.

That said, if we disregard the leaky SDK APIs and half-implemented everything, it does somewhat deliver on the pluggability promise. Before OTel, you had bespoke stacks for everything. Now there is some commonality - you can plug in different logging backends to one standard SDK and expect it to more or less work. Yes, it works less well than a vertically integrated stack but this is still something. It enables competition and evolution piece by piece, without having to replace an observability stack outright (never going to be a convincing proposition).

So while the developer experience is pretty unpleasant and I am also disappointed with the actual daily usage, from an architectural perspective it opens up new opportunities that did not exist before. It is at least a partial win.

While some of the SO practices can feel dumb, I wonder what is the tradeoff we are making here? Could it be that for whatever reason (e.g. personality) we might benefit from accepting these "dumb" choices because it also brings with it unrelated benefits? For example, might it be that if they were forced to accept "frivolous" statements such as "thank you" (from the example you gave), might it cause moderators to not moderate, and thereby allow in spam & other nasty bits?

It is worth bearing in mind that if we, "good people", complain about moderation, we only see the parts of it that touch our "good posting". There might be plenty of good that moderators are also doing, which only the bad guys see.

Sounds like your files are not static, therefore your problem is not the same as the problem the article is addressing.

What is the state of the art for doing encryption with ECC? The author just says "use NaCl" here but what should I do if I am not in a position to do that but can still use ECC?

My understanding of ECC is that it is not really suitable for encryption as-is, as RSA was, rather it is used for key agreement (somehow through a multi-step process that I do not understand). But it is unclear how much of this is just rumor and implementation limitations.

European context here. It is not even an option - you work 40 hours per week, end of story. Working more than 40 hours per week is seen as something wrong, something you need to bring up with a manager to address because someone's work/life balance has gotten off on the wrong path. It is a problem to be addressed.

We have high flexibility in terms of when we work - we can do 4x10 hour days or work a weekend to take some days off in the middle of the week, things like that. But this is an option for the employee, not something the employer can enforce on anyone - the choice is always your own.

If there is a deadline that cannot be met, this means:

1. The company needs to work with the stakeholders to set proper expectations about what will and will not be delivered, perhaps cutting up the work into multiple iterations.

2. The problem should be raised in advance so there is time to address it and prepare the stakeholders to face the facts. If an engineer speaks up on the last day of a 4 month project and says "actually we cannot deliver it, we need one more month" then that is a problem with that engineer's performance (he should have communicated the problem earlier) and/or a lack of leadership (someone should have noticed/cared by then!). It does not change the facts of #1.

This is based on a bit more than 15 years of experience, very consistent from mid-size company to megacorp.

> I've seen people writing C# applications and wondering why stuff is taking 200ms

I observe that in the most recent generation of its HTTP client (SocketsHttpHandler), .NET also sets NoDelay by default.

https://github.com/dotnet/runtime/blob/main/src/libraries/Sy...

I have the unfortunate pleasure of being very sensitive to CO2 levels. At around 800-900 ppm I get a general feeling that something is abnormal. If it goes to 1000 ppm then I can usually tell that I am not thinking well anymore. It is a similar feeling to being slightly drunk - for example, I start a thought but find it difficult to carry it through to the conclusion.

Before I got a CO2 meter, I could always tell within hours when my building's flaky central ventilation fan had failed because I simply felt something was off.

Unless you are similarly cursed/blessed, just get a CO2 meter. I could not imagine learning this, it is like learning to feel temperature.

You sure sound like a driver who owns the road. Nobody expects you to pass on a double yellow. You are supposed to slow down as needed to stay behind the cyclist until it is safe to pass.

Those 3-5 years at Apple could be lucrative for you. Go for it!

I have a vague understanding that it is not so easy to encrypt data with ECC as it is with RSA. Is that true? This is one reason I still use RSA. What is the right way to use an ECC public key to encrypt data so only the holder of the private key can decrypt it? (Without any fancy key exchange - just fire and forget, email style)

Microsoft will be storing EU users' data in the EU: https://blogs.microsoft.com/eupolicy/2021/05/06/eu-data-boun...

Will this enable them to comply with the requirements?

That limitation seems to just be a marketing thing - a few more searches reveal they do not actually check what registrar you use.

M365 is what I am looking at, as well. Can you go into detail about the differences and what you found better with M365?

Does Fastmail do mailing lists? That's the one thing I use in GSuite other than the basic email.

> that can only be used within the e-government services

> You can't use it for other purposes. No RFC defined email address is shared with you

This is not entirely true. You get both:

* idcode@eesti.ee can only be used by government senders.

* you also get first.last.uniqueid@eesti.ee which works as a regular email address.

What exactly is it? I understand there is something about code puzzles but how does it work? How does the leaderboard work - presumably there is some sort of scoring involved?

> it's under the implicit assumption that you or your family will be the only ones occupying it.

Says who? This is a rather bold assumption to make and does not have any connection with the reality in most of the world.

I thought this was the next logical step after Windows started incorporating containers - for several years you saw the two companies get closer together until suddenly they no longer seemed to be on speaking terms. Docker really missed their opportunity here.

> That’s my boring story of entitlement.

Damn straight. Call me a soulless corporate drone but I completely understand AT&T here. Sometimes you need to drop old and unprofitable services.

With millions of customers, you can't afford to cater to everyone who wants to be treated as a special case. There are genuine special cases that need the attention, and "but I like my phone" is just an example of a modern Karen in the wild.

Times change and we must change with them.

A little known feature is that the Docker integration in Visual Studio gives you this automatically.

Sounds like it's time for the game authors to provide some servers with the compute resources to do this background processing.

This has been sitting in some alpha state for years now. Is there any progress being made or is it abandonware?

Microsoft is also working on a kernel-bypass userspace network stack: https://github.com/demikernel/demikernel